Why I built MonkStreet

I bought Amazon in 2013 and I still own it. I bought Nvidia years before most people could pronounce CUDA, and I still own that too. Those two positions, more than anything else I've done in markets, are responsible for the shape of my financial life.

They were qualitative calls — bets on management, on optionality, on what kind of company you'd actually want to work at. They were right for reasons a spreadsheet would have struggled to capture in 2013.

I'm telling you this up front because I don't want you to mistake what follows for an argument against instinct. My instinct has been good to me. Gut is real.

But here's the thing nobody warns you about when they sell you on "just trust your gut": the hard part of a great investment isn't buying it. It's holding it.

Gut, on its own, isn't enough to hold with

When Nvidia dropped 50% while I held it, what kept me in wasn't a feeling. Feelings, at that moment, were doing everything they could to get me to sell.

What kept me in was being able to point at something outside my own head — something that didn't care how scared I was that week — and say: the numbers still look like the numbers of a great business.

That is what I mean by putting a number on your market conviction. It isn't about replacing judgment. It's about giving judgment a floor to stand on when the market is busy trying to shake you off your best position.

The mirror image of that is the reason I had to build MonkStreet at all. I want to tell you about Carvana.

Carvana sold used cars online. You picked one on your phone, and it came to your door, or you collected it from a glass tower that worked like a giant vending machine. It was the kind of idea that makes you feel stupid for not having thought of it. I bought in November 2019, at $79 a share.

Four months later, the pandemic hit and the stock lost almost two-thirds of its value. I held, because the story had just got better: the dealerships were closed, and buying a car from your sofa was suddenly normal.

The story was right, for a while. By August 2021 the stock was around $370, close to five times what I had paid. I told myself I had seen it before everyone else.

In December 2021, with the stock already a third below its top, I bought more, at $232. Most of the money I ever put into Carvana went in that day. Then it kept falling, and I held, because now I was anchored to a price that no longer existed. In May 2022 I sold everything at $30. For every dollar I had put in, I got 18 cents back. Seven months later, the stock was under $4: down 99% from the top.

A long time later — once I'd built the thing I'm about to describe — I went back and ran Carvana through it.

When I first bought it, Carvana scored 0 out of 100. Not a low score. Zero. In the two and a half years I owned it, through the crash, the fivefold rise and the collapse, it never scored above 23. On the day I put in most of my money, it was 22, with Quality at 4 and Safety at 15.

The sales were growing fast. The company was not turning them into profit, it was borrowing to keep going, and the warning signs that often come before a big loss were all over it. For almost two years the price went up anyway. That is the hard part: the numbers can be right and still look wrong for a long time. They had been warning me the whole time, and the rising price was the reason I stopped listening.

The numbers also changed their mind when the company did. By the end of 2023, after Carvana had cut its costs and started to make money, it scored 96. By then I had stopped looking.

Amazon and Carvana were the same kind of bet for me: a story I believed in. With Amazon, the numbers quietly agreed with me through every drawdown, and I could lean on them when my nerve was failing. With Carvana, they disagreed from the first day — and I never checked, because the story was too good to check.

This is what I mean when I say gut needs a floor. Not a cage. A floor. Something underneath you that tells you whether the company you're about to fall in love with looks like the ones that have historically rewarded patience, or the ones that have historically punished it.

When gut and floor agree, you get to hold through a 50% drawdown and come out the other side. When they disagree, you at least get to notice — instead of spending eighteen months defending a thesis to yourself in a voice that sounds suspiciously like hope.

I didn't need a more beautiful story. I needed something to stop me from believing the beautiful ones without checking.

If there's a right answer, go find it

Once I accepted that I needed a quantitative floor, the next question got interesting. What should the floor actually be?

Somewhere in the middle of all this, I read What Works on Wall Street by Jim O'Shaughnessy. It reorganized how I think about investing in a way almost no other book has.

The argument, boiled down: test investment strategies rigorously across decades of real market data, and most of what people tell you about stock-picking turns out to be folklore. A small number of factors — the ones with real statistical backbone — keep working, quietly, across regimes, across decades, across everything.

What hit me hardest was the implication. Qualitative investing is hard to improve because you can't really measure it. Two smart people can read the same 10-K and reach opposite conclusions, and neither is exactly wrong. They're just human.

Quantitative investing is different. It has a right answer. Not about the future — nothing has that — but about what has actually worked, on real data, for real companies, over a window long enough that luck runs out.

If there's a right answer, I wanted to find it. Not a right-ish answer. Not a right-for-my-personality answer. The boring, rigorous, grown-up version: the one that survives proper statistical tests and resists the temptation to overfit.

Because here's the thing about a floor. If you're going to trust it to hold you through a Nvidia-sized drawdown, it had better be solid. A sloppy floor is worse than no floor. A sloppy floor is just a more confident way to be wrong.

That became the project.

What MonkStreet is

MonkStreet is the floor I wish I'd had in 2013, and in 2016, and on the day I bought Carvana.

It's a research platform built around a single number we call the MonkScore. It runs from 0 to 100, and it answers one question: how strong are this company's fundamentals, right now, next to its peers?

Under the hood it's financial factors grouped into five pillars — Growth, Profitability, Quality, Conviction, and Safety — tested across 27 years and 365,193 point-in-time company-quarters from 12,105 US-listed companies above our size floor, with every delisted and bankrupt company kept in. We attacked it with random copies of its pillars, factor controls, and cleaner data — and when the cleaner data made the numbers smaller, we published the smaller numbers. The full methodology — with the statistical tests, the charts, and the honest caveats — has its own page. This isn't that page.

What I want you to take from this one is simpler. MonkScore isn't a prediction. It's a mirror. It holds a company up next to its peers on five pillars at once: this one is strong across the board, this one is weak, this one is somewhere in between.

The Carvana I bought scored 0 out of 100 on the day I first bought it, and never more than 23 while I owned it. I'd still have had to decide what to do with that. But I couldn't have pretended the numbers agreed with the story.

Who it's for

I didn't build MonkStreet for a trading desk inside a bank. I built it for people who do their own thinking — and don't have a quant team behind them to do it for them.

The clearest one in my head is a friend from school. Call him Oscar. He's doing well: two kids, a real job, a calendar with no room to read twenty annual reports before dinner. He genuinely loves investing — he'd read 10-Ks all weekend if the world let him — but it doesn't, and most of his ideas die not because they were bad, but because he never found the time to check them properly.

Oscar was the first person to use MonkStreet, and he said the thing that told me I'd built the right product:

"In a few seconds I can tell whether a stock is worth my research time, or whether I should drop it and move on."

That was the whole thing. Not "tell me what to buy" — he's a grown-up, he wants to make his own decisions. Just: stop me from burning hours on companies the data already says aren't worth it.

The independent advisor is the same person with clients. If you run your own book and want research that stands up to scrutiny, without the infrastructure of a giant firm behind you, MonkStreet is built for you too. Same need: a rigorous, repeatable read you can defend, on every name, the same way every time.

One honest caveat. MonkStreet doesn't predict the future. Nothing does. A MonkScore of 100 today is a company that is strong on all five pillars at once, next to its peers — a real statement about the present, not a promise about next year. The score is a starting point. It tells you where to look and where not to waste your time. The actual work — what to own, at what size, for how long, on what thesis — is still yours. It should be. That part is what makes you an investor.

Come in

Go to app.monk.st and try it. Look up a stock you own, or one you've been thinking about, and see what the score says.

If it agrees with you, good — the numbers are on your side. If it disagrees, even better. That's the conversation I want MonkStreet to start.

The full methodology — backtests, statistical tests, honest caveats — is there whenever you want to dig in. And if you just want the math without an account, our free tools — WACC, dividends, position sizing — are open to everyone.

Until then, happy investing.

— Alberto (The Investing Monk)